312 SEWAGE AND ITS PURIFICATION 



the Manchester observations in Chap. XL I have repeatedly 

 observed storm-water to be even more impure than the ordinary 

 sewage; a sample I analysed in 1905 contained in parts per 

 100,000, free ammonia i*6i, albuminoid '40, oxygen consumed 

 6'0i, chlorine 5*4, nitrogen as nitrates and nitrites '197 ; 

 whereas the sewage gave generally better results, and the land 

 effluent from the same works yielded an average of "6 free 

 ammonia, '12 albuminoid, 053 oxygen consumed, and '57 of 

 oxidized nitrogen. The Royal Commission on Sewage also 

 found,^ that storm-water was almost invariably impure, both 

 chemically and biologically, that street washings were all 

 impure biologically, even when they contained only a small 

 quantity of organic matter, and that the liquid might even be 

 very impure after long continued rain ; and remarked that " the 

 practical advantages of the separate system may be great, and 

 doubtless storm overflows are necessary, but the fact that storm 

 liquids may be so impure, both chemically and bacteriologically, 

 is a point of considerable importance." 



For the safety of the sewers and the avoidance of flooding of 

 basements, it is always necessary to construct storm overflows. 

 Without these in a sewage farm scheme the ground is liable to 

 become waterlogged, and in a filtration process the excess of 

 water by its volume and velocity tends to derange the purifica- 

 tion plant, hence it is usually allowed to escape from the sewers 

 by special outlets when above a certain amount, carrying with 

 it a mixture of the unpurified sewage. The combined system 

 also involves the construction and maintenance of sewers very 

 much larger than the volume of the regular flow, in order to 

 provide for occasional contingencies. This greater capacity 

 presents inducements to the disposal of grosser refuse which 

 would not be tolerated in a smaller sewer, and often it is im- 

 possible — except at rare intervals — to properly flush the entire 

 surface of these large channels. 



The " separate system," in which the sewage proper is kept 

 apart from rainfall and storm-water, has conduits of such size 

 only as to preclude the possibility of the sewage becoming stag- 

 nant therein, the size being governed by the bore of the water 

 main, since if a given diameter of pipe supplies all the water 

 needed, a little above the same diameter should be sufficient for 

 an exit. Mr. Silcock proposes that the rainfall sewers of a 

 separate system should be provided with leaping weirs dis- 



1 1904 Report, vol. iv., part i., p. 105. 



